The cost of cancer to employers is high. Since 44% of those who develop cancer are employed, worksites would be an appropriate location to provide cancer detection and prevention information. Yet prevention programs exist in worksites today. Dietary changes and eariy datection scmening could de the incidence of cancer an estimated 25% to 50%. It is still to be determined if the workplace is an effective site for the delivery of nutrition and screening education and behavior change interventions specifically designed to increase cancer prevention nutrition and screening behaviors. Community behavior change models have been used for planning community centered lifestyle intervention projects with some success. Because the worksite can be considered a community with many of the same interpersonal and organizational factors operating, although on a smaller scale, the community behavior change model will provide the conceptual framework for the worksite intervention research planned. Participants will consist of GTE employees recruited from 11 matched worksites in the Tampa, Florida area. Sites have on matched on existence of food service, size, job classifications, and gender. The total number of employees is 2500, but the unit of analysis is the worksite. GTE top management and labor have enthusiasticaUy endorsed the GTE/University of Florida partnership for the purpose of conducting this research. A major grocery chain has agreed to become involved in offering educational opportunities, in addition to providing various donations. Intensive interventions will include education and action plan groups emphasizing problem solving techniques, self-instructional materials, and cafeteria modification, Incentives, games, lotteries, coupons and a favorable pricing strategy will be key factors in encouraging participation and behavior trial. Social marketing strategies will be used to facilitate strategic planning and to involve participants in programmming and formativ evaluation to allow mid-course correction. A pre post test design, utilizing matchcd pairs of worksites, randomly assigned to treatment or control groups will be used. A computer-assisted food frequency program has been identified that can be modified to include questions on cancer risk, in addidon to diet and related cancer knowledge and attitudes. Evaluation will consist of comparisons of differences betwee experimental and control groups.